Highlights of Recent Epistemology

7/27/2019 Highlights of Recent Epistemology

http://slidepdf.com/reader/full/highlights-of-recent-epistemology 1/30

S U R V E Y A R T I C L E

Highlights of Recent EpistemologyJames Pryor

ABSTRACT

This article surveys work in epistemology since the mid-1980s. It focuses on (i)

contextualism about knowledge attributions, (ii) modest forms of foundationalism,

and (iii) the internalism/externalism debate and its connections to the ethics of belief.

1 Contextualism

1.1 Contextualism and relevant alternatives theories

1.2 Ruling an alternative out

2 Modest forms of foundationalism

3 The internalism/externalism debate

3.1 Simple Internalism and Access Internalism

3.2 Internalism about inferences and grounds

3.3 A spectrum of views

4 The ethics of belief 4.1 What does it mean to say that justi®cation is `normative'?

4.2 Justi®cation, warrant, and epistemic blamelessness

4.3 How should one's ethics of belief aect one's views about the internalism/

externalism debate?

The period from the mid-1970s through the mid-1980s was a period of bold

new positions in epistemology. We can group the major developments of that

period under three headings: (i) relevant alternatives theories of knowledgecame on the scene; (ii) foundationalism made a come-back, and the debate

between foundationalists and coherentists intensi®ed; and (iii) we saw the

development of externalist theories of knowledge and justi®cation, most

prominently Goldman's reliabilist account of justi®cation ([1979], [1986], Chs

4±5) and Nozick's account of knowledge ([1981], Ch. 3).1

This paper surveys what has been happening in epistemology since those

developments. I will not attempt to address all the important recent work in

epistemology; but will focus on the issues that seem to me to have occupied

Brit. J. Phil. Sci. 52 (2001), 95±124

&British Society for the Philosophy of Science 2001

1 I won't be able to discuss Nozick in this survey, but his account of knowledge has been an

important topic during the past twenty years. (See Forbes [1984]; McGinn [1984]; Goldman

[1986], Ch. 3; Luper-Foy [1987]; Plantinga [1988]; Fumerton [1995], Ch. 4; and DeRose [1995]

for some of the critical literature.) Kripke has a well-known discussion of Nozick's account of

knowledge, but unfortunately this remains unpublished.



center stage. We can group those issues under three headings, too,

corresponding to the three developments cited above: (i) contextualist

theories of knowledge came on the scene and are now starting to be widely

discussed; (ii) modest forms of foundationalism have become more familiar,

and are now widely believed to be viable; and (iii) epistemologists have begun

more vigorously to explore issues in the ethics of belief, and their connection

to the internalism/externalism debate. I will discuss each of these in turn, and

will devote the most attention to issue (iii).

I should note one important omission: I will say little about `naturalized

epistemology' in this survey, though that has received a good deal of

attention in recent years. In part, this is because it is hard to ®nd a single set

of theses one can claim to be de®nitive of naturalism. But largely it is because

of space constraints. Since naturalism has been well-surveyed elsewhere (seeMae [1990b]; Kitcher [1992]; Goldman [1994]; and Kornblith [1995]), I

thought it would be better to concentrate on the issues I do take up below.2

1 Contextualism

1.1 Contextualism and relevant alternatives theories

We'll understand Contextualism to be the doctrine that the truth of aknowledge-ascription can vary with the context in which that ascription is

made or assessed. For example, according to the Contextualist, the sentence

`Reagan knew that the Iran/Contra dealings were taking place' might express

a hard-earned truth when uttered in a history classroom, yet express

something false when uttered in a discussion of skepticism.3

Contextualism grew from a number of dierent roots. Firstly, it was

common in discussions of skepticism to invoke a distinction between the high

standards for knowledge that the skeptic seems to demand and the lower

standards for knowledge that seem to be in place when we're making

96 James Pryor

2 Other issues that have been prominent recently, but which I lack the space to discuss, include:

renewed interest in the topic of a priori justi®cation (see fn. 14 for references); Sosa's work on

virtue epistemology (see [1991]) and the development of this ®eld (for surveys, see Axtell [1997];

Greco [1993], [1999]); how externalism about content aects one's capacities for self-knowledge

(this is discussed in some recent collections: Ludlow and Martin [1998]; Wright et al. [1998];

Boghossian and Peacocke [2001]); and a greater attention to testimony and to social aspects of

knowledge more generally (for an overview, see Fuller [1996] and Schmitt [1999]).3 For recent defenses of Contextualism, see: Unger ([1986]); Cohen ([1987], [1988], [1999]); Heller

([1989], [1999a]); DeRose ([1992], [1995]); and Lewis ([1996]). For criticism, see Schier ([1996]);

Feldman ([1999a]); and Vogel ([1999]). DeRose ([1999]) gives a useful overview of the ®eld.

Contextualism in the sense de®ned here is importantly dierent from the view Annis ([1978]) andWilliams ([1996], [1999b]) defend, which makes a subject's epistemic status depend on facts about

the subject's circumstances: e.g. what epistemic standards his community endorses, what kinds of

activity he's engaged in, and so on. Contextualism makes a subject's epistemic status depend on

facts about the context in which that epistemic status is being appraised . This might be dierent from

the context the subject occupies. See the discussion of RA-Theory without Contextualism, below.



ordinary epistemic evaluations. The Contextualist can allow that both of

these standards are legitimate, in their proper context. (And the Contextualist

is not limited to just two kinds of standards: there may be many dierent

standards, appropriate to dierent contexts.)

Another early root of Contextualism was the discussion of Gettier-cases

where evidence one does not possessÐe.g. a newspaper headline one hasn't

read yetÐseems to defeat one's claims to knowledge (see Harman [1968],

[1973], Ch. 9, [1980]).4 It is very natural to think that the extent to which this

kind of evidence aects one's epistemic standing will vary with context.

Suppose your friend in the White House tells you that the President is in

Mexico, and suppose you truly believe on that basis that the President is in

Mexico. But suppose also that, unbeknownst to you, the Washington Post has

just printed a (false) story that claims that the President is not in Mexico butrather vacationing in Maine. In some contexts we might want to count your

justi®ed true belief as knowledge despite the existence of the misleading

newspaper report. In other contexts, though, we'd take the fact that the

newspaper report is widely disseminated to preclude your belief about the

President's whereabouts from counting as knowledge. The Contextualist can

make sense of these shifts.

A third root of Contextualism is the Relevant Alternatives Theory of

knowledge put forward in the 1970s by Dretske ([1970], [1981]) and Goldman

([1976]). We should discuss this at length.

As we'll understand the RA-Theory, it says that, if q is an irrelevant

alternative to p, then knowing p doesn't require you to have evidence which

would enable you to rule q out; in fact, so long as q is irrelevant, you don't

need to have any independent evidence against q. For example, if you see a

zebra in a zoo pen, and the possibility that the animal you see is a cleverly

disguised mule is an irrelevant alternative, then you would not need to have

evidence against that alternative, in order to know that the animal is a zebra.

This RA-Theory combines naturally with Contextualism. The combinedview says that what alternatives count as relevant is a function of the context

in which the knowledge ascription is made or assessed (see Stine [1976];

DeRose [1992], [1995]; and Lewis [1996]).

Now, early proponents of the RA-Theory, like Dretske, denied that

knowledge is closed under known logical consequence. But contemporary

Contextualist proponents of RA-Theory argue that we should retain some

kind of Closure Principle. Let q be some alternative you cannot rule out.

Their idea is that, in a given context where q is irrelevant, then if you know p

and you know that p implies not-q, you count as knowing not-q, too. Some

Highlights of Recent Epistemology 97

4 Cohen ([1987]) discusses a similar phenomenon: the extent to which evidence one does possess

but doesn't appreciate the force of can defeat one's claims to knowledge.



contextualists say that when we explicitly raise the question whether you know

not-q, that makes q relevant, and hence, since you can't rule q out, you don't

count as knowing not-q, and you don't count as knowing p either. But as they

see it, there is no single context in which you both know p and fail to know

not-q. In contexts where q is relevant, you count as knowing neither. In

contexts where q is irrelevant, you count as knowing both.5

Although RA-Theory is nowadays often combined with Contextualism, it

is important to recognize that these two theories don't have to be combined.

Some RA-Theorists believe that what alternatives are relevant is wholly

determined by facts about the putative knower and his environment (perhaps

including facts about his epistemic community). It does not vary from one

context of knowledge-ascription to another. That is a form of RA-Theory

without Contextualism.6

One can also hold Contextualism without RA-Theory. This is the best way

to understand Cohen ([1998], [1999]). Let's call this view Strength-of-Evidence

Contextualism. (Vogel [1999] calls it `Plain Contextualism'.) The Strength-of-

Evidence Contextualist thinks that what you count as knowing varies from

context to context, but he denies that you can ever, in any context, know a

proposition while having no evidence against alternatives to that proposition.

Facts about the context in which we're ascribing knowledge to a subject never

excuse the subject from the requirement to have some evidence against

alternatives to what he believes. For instance, the Strength-of-Evidence

Contextualist says that until you have at least some evidence against the

possibility that the animal you see is a cleverly disguised mule, you can't know

that animal to be a zebraÐeven if the possibility that the animal is a cleverly

disguised mule is an irrelevant alternative. What context does aect is how

strong your evidence has to be, what newspapers you have to take into account,

what defeating evidence you need to appreciate the force of, and so on. In

some contexts, you can know that the animal you see is a zebra, while having

only very slender evidence against the possibility that it's a cleverly disguisedmule; in other contexts, you need a lot more evidence. By the same lights, the

Strength-of-Evidence Contextualist says that in order to count as knowing

anything about the external world on the basis of perception, you'd need to

have at least some reasons for rejecting hypotheses about brains in vats and the

98 James Pryor

5 Stine ([1976]); Cohen ([1988]); DeRose ([1995]); and Lewis ([1996]) all argue for Contextualist

views which retain some form of Closure. Heller ([1999b]) defends Dretske's rejection of

Closure. For further discussion of the relations between Contextualism and Closure, see Vogel

([1990]).6 This appears to be Dretske's view ([1991]). It is also formulated by Goldman ([1976]), though

Goldman doesn't endorse the view. He prefers a Contextualist version of RA-Theory. On the

dierence between RA-Theory with Contextualism and RA-Theory without Contextualism,

see Cohen ([1991]) and DeRose ([1992], [1999]). DeRose calls RA-Theory without

Contextualism `subject contextualism', contrasting it to the `attributor contextualism' he

advocates.



like, even in contexts where those hypotheses are irrelevant. Strength-of-

Evidence Contextualism does not itself oer any story about where those

reasons come from. We need some further, independent account of that.7

1.2 Ruling an alternative out

In discussions of RA-Theory and Contextualism, most attention has been paid

to the question of what makes an alternative relevant. (See Vogel [1999], section 3

for discussion.) Comparatively little attention has been paid to how we manage

to eliminate or rule out an alternative. So let's say a few words about this issue.

On Dretske's account, ruling q out is just a matter of knowing not-q.

However, this account of ruling out would not suit those RA-Theorists who

wish to retain Closure. For suppose that q is some alternative you can't rule

out. According to the RA-Theorist, you can sometimes know p despite being

unable to rule q outÐso long as q is irrelevant. If q were to become relevant,

then your inability to rule it out would prevent you from knowing p. Now, if

we retain some form of Closure, then in those contexts in which you count as

knowing p you'd also, by Closure, count as able to know not-q. This despite

the fact that you're not able to rule q out. So if we retain Closure, we should

understand ruling an alternative out as requiring more than just knowing that

alternative to be false.A better account might be an evidential account of ruling out. We could say

that ruling q out requires you to have independent evidence against q,

evidence which does not beg any of the questions for which q is an

alternative.8 An RA-Theorist who adopts this account of `ruling out' can say

that, when the possibility that you are a brain in a vat is irrelevant, you can

know by perception that you have hands, and hence by Closure also know

that you're not a handless brain in a vat. But you're not in a position to rule

outthe possibility that you are a brain in a vat, because you have noindependent evidence against that possibility. Whatever justi®cation you have

for believing that you're not a brain in a vat crucially rests on things you


7 As I said, Cohen ([1988], [1999]) is a Strength-of-Evidence Contextualist. He does not try to give

any story about what justi®es our rejection of run-of-the-mill alternatives, like the cleverly

disguised mule. He does, however, think we have some inductive evidence that those alternatives

do not obtain. More problematic are global skeptical hypotheses, like the hypothesis that one is a

brain in a vat. Cohen thinks no possible evidence could tell against these alternatives. But he thinks

they are in some sense a priori irrational, that we're entitled to reject them without evidence

(compare BonJour [1985], sections 8.3±8.4).

Stine, on the other hand, rejects Strength-of-Evidence Contextualism in favor of a Contextualist

version of RA-Theory. She says you can know that the animal in the pen is not a cleverlydisguised mule merely by virtue of the fact that the disguised-mule hypothesis is irrelevant. You do

not need any speci®c evidence that tells against it (Stine [1976], pp. 257±58). Some Contextualists

like Lewis ([1996]) follow Stine here. Others like Heller ([1999b]) follow Cohen.8 See the discussion of antecedent justi®cation in Pryor ([2000]); and the discussion of

`transmission failure' in Wright ([1985], [2000]) and Davies ([1998], [2001]).



purport to know about the external world by perception. It's for this reason

that when the possibility that you are a brain in a vat becomes relevant, you

no longer count as knowing you have hands.

A dierent account construes ruling out in counterfactual terms. According

to this account, you rule an alternative out by having your belief be

counterfactually sensitive to whether that alternative obtains. For example,

when it's your knowledge of p that is in question, such an account could say

that being able to rule q out is a matter of being such that, if q were true, you

would no longer believe p.9

The accounts of ruling out mentioned so far attempt to be explanatory in a

certain way. They appeal to some independent facts about your relation to

your evidence in order to explain facts about what that evidence enables you

to know. An alternative way to proceed would be to take as given the factsabout what your evidence would enable you to know in certain contexts, and

use those facts to explain the notions of `being relevant' and `ruling out'. This

is Cohen's initial approach; he de®nes an alternative to p as `relevant' just in

case it has to be ruled out in order for you to know p ([1988], pp. 100±1).

However, Cohen then goes on to discuss more substantial criteria for being

`relevant'. A similar view would say that you're in a position to rule some

alternative to p out just in case you'd be able to know p even if that

alternative were relevant. These construals of `relevant' and `ruling out' are

perfectly legitimate, as far as they go. They just won't be very useful for

giving explanations of what we know when. They take the facts about what

we know when for granted, and ride piggy-back on them.

Though this issue of how to understand the notion of `ruling out' needs more

attention, Contextualism is currently a very active research program, and it has

rapidly become one of the most seriously regarded responses to skepticism.

2 Modest forms of foundationalismAnother active source of resistance to skepticism comes from new

foundationalist accounts of perceptual justi®cation. Foundationalism's

prospects quite generally look much better today than they did twenty-®ve

years ago. In this section, we'll take a brief look at this development.

Say that you are `immediately justi®ed' in believing p i you're justi®ed in

believing p, and this justi®cation doesn't rest on justi®cation you have for

believing other supporting propositions. According to the foundationalist,

100 James Pryor

9 This proposal is akin to Nozick's account of knowledge; however, the current proposal onlyrequires your beliefs to be counterfactually sensitive to the obtaining of relevant alternatives.

Goldman ([1976]); DeRose ([1995]); and Heller ([1989], [1999a], [1999b]) favor this account of

ruling out. A related view says that you are in a position to rule q out i, if q were true, you

would not have the same evidence you in fact have for believing p. Lewis ([1996]) employs an

account similar to this.



the justi®cation for all our beliefs ultimately traces back to a set of `basic

beliefs', which we have immediate justi®cation for believing.

Traditionally, foundationalists thought that these basic beliefs could only

concern the nature of one's current thoughts and experiences. (And perhaps

some a priori matters as well.) All our other empirical beliefs, including beliefs

about our perceptual environment, had to rest inferentially on this austere

foundation of beliefs about our current mental states. These basic beliefs were

thought to be exceptionally secure. They were often claimed to be infallible

and indubitable. According to many foundationalists, the reason why these

beliefs were so secure was that we had a non-propositional `direct

apprehension' of the mental states the beliefs were about.

There are many reasons to be unhappy with that traditional cluster of

views. It's generally felt that a foundation consisting only of infallible beliefsabout our own mental states would be too austere to support the rest of our

empirical beliefs. And many philosophers also criticized the traditional

foundationalists' views about `direct apprehension'.10

In addition, BonJour put forward an in¯uential argument which seemed to

show that, at least on internalist conceptions of justi®cation, it was impossible

for there to be any basic beliefs. BonJour defended an account of justi®cation

according to which, to be justi®ed in believing a proposition p, you have to be

justi®ed in believing that your belief has some features which make it likely to be

true. This account makes it impossible for your belief in p ever to be immediately

justi®ed . Your justi®cation for your belief in p would always rest in part on your

justi®cation for those further beliefs about the belief's features.11


10 See the anti-Given arguments in Sellars ([1956]); Davidson ([1981]); BonJour ([1978], [1985]);

and Williams ([1999a]). BonJour has changed his mind about these arguments; see below.

These anti-Given arguments deserve a re-examination, in light of recent developments in the

philosophy of mind. The anti-Given arguments pose a dilemma: either (i) direct apprehension is

not a state with propositional content, in which case it's argued to be incapable of providing us

with justi®cation for believing any speci®c proposition; or (ii) direct apprehension is a state

with propositional content. This second option is often thought to entail that direct

apprehension is a kind of believing, and hence itself would need justi®cation. But it ought

nowadays to be very doubtful that the second option does entail such things. These days many

philosophers of mind construe perceptual experience as a state with propositional content, even

though experience is distinct from, and cannot be reduced to, any kind of belief. Your

experiences represent the world to you as being a certain way, and the way they represent the

world as being is their propositional content. Now, surely, it's looking to you as if the world is a

certain way is not a kind of state for which you need any justi®cation. Hence, this construal of

perceptual experience seems to block the step from `has propositional content' to `needs

justi®cation'. Of course, what are `apprehended' by perceptual experiences are facts about your

perceptual environment, rather than facts about your current mental states. But it should at

least be clear that the second horn of the anti-Given argument needs more argument than we've

seen so far. To be fair, most of the foundationalists who appealed to a notion of `direct

apprehension' understood this notion along the lines of horn (i); and that is where their criticshave focused most of their attention.

11 See BonJour ([1978], [1985], Ch. 2). BonJour's account of justi®cation is a version of what I call

`Access Internalism' in Section 3.1, below. BonJour attempts to motivate this account by

appeal to considerations about epistemic responsibility. We will discuss the connection between

justi®cation and notions like epistemic responsibility in Section 4, below.



In recent years, though, foundationalism has staged quite a successful

comeback.12 There are two main reasons for this. In the ®rst place,

foundationalists nowadays think that the implausible elements in tradi-

tional foundationalism are inessential extras that aren't required by the

reasoning that primarily motivates foundationalism. Newer forms of

foundationalism are quite `modest', in that they allow basic beliefs to be

fallible, revisable, less than maximally justi®ed, and so on. They only

require that these basic beliefs be immediately justi®edÐthat is, justi®ed in

a way that doesn't rest on justi®cation one has for other, supporting

propositions. Some of these new foundationalist views allow beliefs about

our perceptual environment to qualify as basic. Some of them even grant

important epistemic roles to facts about coherence: e.g. they may allow

facts about coherence to defeat or to strengthen one's justi®cation for abelief, even a basic belief. These theories count as foundationalist because

they say that, for every justi®ed belief a subject has, at least some of his

justi®cation for that belief is immediate or ultimately rests on other

justi®cation that is immediate.

In addition, many philosophers have come to think that BonJour-style

arguments against the possibility of basic beliefs are unsuccessful. The

now-standard reply to BonJour's argument is that it builds into justi®ed

beliefs about the world conditions which really ought only apply to our

higher-level, re¯ective beliefs that our beliefs about the world are justi®ed .

For you simply to have a justi®ed belief about the world, this reply says,

that belief may need to have features that make it reasonable or likely to

be true, but you need not, in addition, be justi®ed in believing it has such

features. You can have a justi®ed belief `There is a hand', without also

being justi®ed in believing `I am justi®ed in believing that there is a hand'

or `I believe that there is a hand on grounds which make that belief likely

to be true'. Acquiring justi®cation for the latter beliefs would be a further

epistemological accomplishment. Once this move is made, it is hard forBonJour's argument to get going.13

A notable recent development is that BonJour, who was formerly a

leading coherentist about empirical beliefs, has given up coherentism and

102 James Pryor

12 For some recent defenses of foundationalism, see Alston ([1989], Essays 1-3); Pollock and Cruz

([1999], Ch. 5, Section 4.2 and Ch. 7, Sections 1±3); and Audi ([1993], esp. Chs 3, 4, 10, and 12).

I advocate a form of modest foundationalism about perceptual beliefs in Pryor ([2000]).

McGrew ([1995]) argues against modest foundationalism, in favor of a more traditional view

with infallible foundations. Recent years have also seen a number of attempts to revive the

notion of direct apprehension. See Fumerton ([1985], [1995]); Moser ([1985], [1989]); Fales([1996]); and BonJour ([1999b]). Finally, one might count Foley ([1987]) and Haack ([1993]) as

foundationalists, though their views dier in important ways from the ones just cited. A good

survey of foundationalist work in the late 1970s and 1980s is Triplett ([1990]).13 This is emphasized in various articles in Alston ([1989]), esp. Essay 6; and in Audi ([1993]), esp.

Chs 4 and 11.



now himself advocates a form of foundationalism (see BonJour [1999a],

[1999b]).14

3 The internalism/externalism debate

3.1 Simple Internalism and Access Internalism

Let's turn from debates about the structure of justi®cation to debates about

whether justi®cation is internalist or externalist. First, let's ask what we mean

when we call an account of justi®cation ìnternalist'.15

Internalism in epistemology is a dierent sort of issue than internalism in

philosophy of mind. Philosophers of mind are concerned with whether

certain mental properties (e.g. the property of having a belief with such-and-

such content) supervene on one's intrinsic make-up. Epistemologists are more

concerned with the connection between one's justi®catory status and facts to

which one has a special kind of access Ðe.g. facts about what one believes,

what experiences and memories one has, facts about what one's beliefs are

based on (see below), what one's goals are, what one is attempting to do, and

so on. Clearly, not every state that is ìnternal' in the philosophy of mind

sense will be among these states. Sub-personal states of one's visual

processing system are ìnternal' in the philosophy of mind sense, but theyare not states to which we have any special ®rst-person access. In addition, it

is a hotly debated question what sorts of special ®rst-person access we have to

the `wide' contents of our thoughts (see the collections cited in fn. 2). So it is

not obvious that every state which is ìnternal' in the epistemologist's sense

will be ìnternal' in the philosophy of mind sense. Hence, it is certainly

possible to combine externalism in philosophy of mind with internalism in

epistemology, and vice versa.

What kind of `special access' is at play in the debate about internalism in

epistemology? Dierent accounts of this are possible. One could understand this

`special access' in terms of the strength of one's access: one could understand it as

meaning infallibility, say, or as meaning that non-culpable mistakes are

impossible. Alternatively, one could understand this `special access' in terms of

the route by which one has access: one could understand it as meaning that one


14 BonJour has also been at the forefront of another recent movement, which is the development of

modest accounts of a priori knowledge, and a revival of interest in rationalism (see BonJour [1998];

see also Edidin [1984]; Casullo [1988]; Burge [1993]; Plantinga [1993b], Ch. 6). These accounts of a

priori knowledge often incorporate many elements of the new modest forms of foundationalism.

George Bealer has written a series of papers on these issues, starting with Bealer ([1987]), andsummarized in Bealer ([1999]). These will be incorporated in a forthcoming book. Two useful

collections, including some of the works just cited, as well as other important papers, are Moser

([1987]) and Casullo ([1999]).15 For other discussions of this question, see Alston ([1989], Essay 8); Fumerton ([1988], [1995],

Ch. 3); and Feldman and Conee ([forthcoming]).



can know by re¯ection alone whether one is in one of the relevant states. (By

`re ection' I mean a priori reasoning, introspective awareness of one's own mental

states, and one's memory of knowledge acquired in those ways.) It would then be

a further question whether one's re¯ective knowledge was infallible or especially

certain in any remarkable way. Dierent choices about how to understand the

notion of `special access' will aect one's understanding of the entire internalism/

externalism debate. Most epistemologists understand the notion in the last way I

described, and I will follow them.

Now, there are two dominant ways in which epistemologists understand

the notion of ìnternalism'. One is in terms of a supervenience thesis. The

other is in terms of whether the subject has special access to her epistemic

condition. We'll start with the supervenience construal of ìnternalism'.

I will take the following supervenience thesis to express the most minimalinternalist position:

Simple Internalism: Whether one is justi®ed in believing p supervenes on

facts which one is in a position to know about by re¯ection alone.

I will take Externalism to be the denial of Simple Internalism.

Now, sometimes when we're evaluating your epistemic standing, we're only

interested in whether you have justi®cation for believing certain proposi-

tions Ðregardless of whether you actually do believe those propositions.

Other times, though, we're interested in more than that. We want to know

whether you do believe the propositions you have justi®cation for believing,

and if so, whether your belief is based on that justi®cation. You may have

very good reasons for believing p, but base your belief in p on bad reasons. In

such cases, your belief is epistemically defective, even though it's a belief in a

proposition you have justi®cation for believing.16

The internalism/externalism debate bears on both of these kinds of epistemic

evaluation. We can ask whether the facts that you're in a position to know

about by re¯ection alone determine that you are justi®ed in believing some proposition. And we can also ask whether those facts determine that your belief

is based on something that justi®es it. One normally sees externalist positions

formulated in terms of the second kind of evaluation. For example, process

reliabilism is usually formulated in terms of whether the process which led to

104 James Pryor

16 In such cases, we say that your belief is ìll-founded', though it's a belief in a proposition you

have good reason to believe. See Feldman and Conee ([1985]) on the contrast between `well-

founded' and ìll-founded' beliefs. A belief counts as well-founded i the subject has grounds

that support the proposition he believes, his belief is based on those grounds, and the subject

isn't ignoring any evidence he has which defeats those grounds. See also Audi's discussion of

the contrast between ìmpersonal' and `personal' justi®cation, and the contrast between reasonsone has for a belief and the reasons for which one holds a belief (in Audi [1993], Chs 3, 7, and 8).

Korcz ([1997]) is a useful survey of literature on the basing relation.

Sometimes one will in forming a belief base it on one ground, but then later come to base the

belief on another ground. This later ground will then be what sustains one's continued

acceptance of the belief. For our purposes we can ignore such complications.



one's actually forming a given belief is reliable. This makes it look like the

debate between process reliabilists and their opponents solely concerns the

evaluation of beliefs. But Goldman pointed out that the internalism/externalism

debate can be conducted in terms of either sort of evaluation (see his discussion

of ex post vs. ex ante justi®cation in Goldman [1979]). The focus in these

discussions on the reliability of our belief-forming processes is sometimes just an

indirect way of raising questions about the reliability of certain sources of

justi®cation, like our senses. Consider a community whose visual experiences

are reliable guides to their environment, but who have a cognitive defect that

leads to them never basing any beliefs on their visual experiences. One can

imagine a reliabilist saying that these people's experiences give them justi®cation

for believing certain things, even though they have no well-founded visual beliefs.

That would be a reliabilist view about the evaluation of propositions, ratherthan the evaluation of actual beliefs.

Whichever sort of evaluation we focus on, we should be careful to

distinguish Simple Internalism from the following, more demanding

internalist position:

Access Internalism: One always has `special access' to one's justi®catory

status.

A strong version of Access Internalism would say that whenever you're

justi®ed in believing p, then you're justi®ed in believing that you are so

justi®ed. This is akin to the `KK Principle', which says that whenever you

know p, you also know that you know. (Nozick [1981], p. 281 construes

`internalism' as the acceptance of such a KK Principle.) Weaker versions of

Access Internalism are also possible. For instance, one might claim that

whenever you're justi®ed in believing p, then the fact that you are so justi®ed

is one you're in a position to acquire justi®cation for believing, by a certain

route. This does not require that you actually already be justi®ed in believing

those facts. Like Simple Internalism, Access Internalism can be formulatedeither as a view concerning the evaluation of propositions or as a view

concerning the evaluation of beliefs.17

`Access Internalism' is now a somewhat-established name for the view(s) just

described. However in some ways this nomenclature is unfortunate, for both

Simple Internalism and Access Internalism have something to do with

considerations of access. The Simple Internalist says that the accessible facts

form a supervenience base for whether we're justi®ed; the Access Internalist says

that whether we're justi®ed is itself an accessible fact. It is important to

recognize that Simple Internalism does not by itself entail that stronger claim.

Even if the fact that one is justi®ed is wholly determined by facts of which one is


17 Chisholm and BonJour (in his coherentist period) were paradigm Access Internalists. (See e.g.

Chisholm [1977], Ch. 6, Section 5; BonJour [1985], Chs 1±2.)



re¯ectively aware, that does not guarantee that one can tell just by re¯ection

that one is justi®ed. After all, one might be epistemologically untrained, or,

worse, have false beliefs about what it takes to be justi®ed. There's no guarantee

that one would be able to correct these defects by mere armchair re¯ection.18

The way I formulated Access Internalism, it imposes an access requirement

on all of one's justi®cation. One might feel that this is too strong, especially

for basic, unre¯ective kinds of justi®cation like we get in perception. Access

Internalism seems to say that our perceptual beliefs are justi®ed only if we're

justi®ed in believing that they are justi®ed. But whether and how our

perceptual beliefs are justi®ed is a matter that epistemologists are constantly

investigating. It seems harsh to say that no one's perceptual beliefs are

justi®ed unless he has a justi®ed position on that matter. In light of this, some

philosophers prefer to impose access requirements only on the kind of justi®cation we seek from inferences. Other philosophers require us to have

access only to the presence of the stu that justi®es our beliefs (e.g. our

experiences), and not to the fact that our beliefs are justi®ed by that stu.

Let's look at these possibilities brie¯y.

3.2 Internalism about inferences and grounds

When you base a belief on some (genuine or alleged) body of supporting

justi®cation, call that justi®cation your `grounds' for your belief. Dierent

things might count as grounds. In some cases, you will believe p on the basis

of other beliefs; these other beliefs are inferential grounds for your belief in p.

When you base your belief in p on an inferential ground, say the belief q, it is

widely agreed that your belief in p is justi®ed only when your belief in q is also

justi®ed. But there is disagreement about what more is required for your

belief in p to be justi®ed. (i) Do you need to have good reason to believe that

your ground q constitutes good evidence for p? Or (ii) does q merely have to

in fact be good evidence for p? Call someone who arms (i) an Inferential

Internalist. An Inferential Externalist denies (i) and arms (ii) instead.19

106 James Pryor

18 If we understand the relevant kind of `special access' in more demanding ways, it will be even

clearer that simple forms of internalism don't entail Access Internalism. For example, the fact

that one's justi®catory status is determined by facts of which one is currently conscious, or

about which one is infallible, does not entail that one is currently conscious of, or infallible

about, what one's justi®catory status is.19 I borrow these terms from Fumerton ([1995], esp. Ch. 3). Fumerton defends something like the

view I'm calling `Inferential Internalism' (but see Cohen [1998] for some important distinctions

which I'm glossing over here). See also the discussion of the `connecting belief' requirement inAudi ([1993], Ch. 8). Feldman and Conee ([forthcoming]) are Simple Internalists who endorse

Inferential Externalism.

Some Inferential Internalists require not merely that q is justi®ably believed to be, but also

that q is in fact, good evidence for p. Their commitment to (i) is what makes them Inferential

Internalists.



Inferential Internalism is a kind of Access Internalism, limited just to one's

inferential justi®cation. (It says nothing about what kind of access we have to

our immediate justi®cation.) Denying Inferential Internalism does not force

one also to deny Simple Internalism. Whether an Inferential Externalist

accepts or rejects Simple Internalism will depend on what he thinks is

required for q to be good evidence for p. The Simple Internalist will argue

that whether q is good evidence for p is a necessary matter, or at least, that it's

wholly determined by facts to which one has the relevant kind of `special

access'; the externalist will argue that it depends on contingent facts about

one's environment, to which one has no `special access'. These positions are

both compatible with Inferential Externalism.

We've been discussing cases where your belief in p is based on other beliefs.

In other cases, you will believe p but not on the basis of other beliefs; rather,your justi®cation for believing p will come from the fact that you have certain

experiences or memories, or from the fact that you understand certain

propositions, or from some a priori intuition. These sorts of things are

immediately justifying grounds for your belief in p. We have to proceed

carefully here. To say that certain experiences are grounds for your belief

does not mean that your belief is based on further beliefs about those

experiences. Rather, on the views I'm envisaging, merely having those

experiences gives you immediate justi®cation for believing certain thingsÐ

e.g. for believing that there is a table. When that's so, you needn't base your

table-beliefs on any beliefs about your experiences.20 However, we still need to

distinguish between subjects who base their table-beliefs on the immediate

justi®cation their experiences aord them, and subjects who, despite having

that experiential justi®cation, none the less base their table-beliefs on other

evidence (e.g. the stories their grandmother told them about tables). That is

why we need the notion of basing a belief on an immediately justifying

ground. Philosophers who deny that any beliefs are immediately justi®ed will

of course deny that a belief can be justi®ed by virtue of being based on animmediately justifying ground.

Even when we count things like experiences as a kind of ground, it is still

controversial whether justi®ed beliefs are always based on some ground. For

instance, it's unclear whether self-evident beliefs and some types of memorial

belief are based on any grounds (see Plantinga [1993b], pp. 187.). But let's

restrict our attention to beliefs that are based on grounds. We can ask several

questions about the access we have to those grounds: Must a ground be


20 I defend such a view in Pryor ([2000]). A more common view is that one's justi®cation for beliefsabout tables derives from one's justi®cation for beliefs about one's experiences. On that view,

one's experiences only serve as immediately justifying grounds for beliefs about one's experiences,

and not for any beliefs about external objects. Acceptance of the view that there are immediately

justifying grounds need not, though it may, be combined with the view that you have any kind of

`direct apprehension' of the ground, from which your justi®cation for the belief derives.



something to whose presence one has the `special access' which is operative in

the internalism/externalism debate? Must one also have special access to

whether the ground is an adequate one, that is, to whether it's possible to

justi®ably base a belief on that ground? Endorsing the weaker access

requirementÐnamely, that grounds have to be things to whose presence one

has special accessÐdoes not by itself commit one to Simple Internalism. For

it could be that one tells an externalist story about what makes the ground

adequate.21 Alternatively, one could tell a Simple Internalist story about why

a given ground is adequate, but refrain from claiming that all subjects will be

in a position to know that the ground is adequate. That would be Simple

Internalism without Access Internalism. If one endorsed the strongest access

requirementÐnamely, that subjects must have special access both to the

presence and to the adequacy of their groundsÐthen one would be a full-¯edged Access Internalist (at least, concerning beliefs which have grounds).22

3.3 A spectrum of views

With these distinctions in hand, we can sketch out a spectrum of internalist

and externalist views.

(i) The strongest form of internalism advocates Access Internalism

about all of one's beliefs. (BonJour in his coherentist period is an

example of someone who held this view.)

(ii) An intermediate form of internalism holds Inferential Internalism,

but denies that one needs special access to one's immediate

justi®cation. (Fumerton holds this view.)

(iii) The weakest form of internalism eschews all forms of Access

Internalism. On such a view, the fact that one's grounds are adequate

must be an internally determined aair, but one need not have special

access to whether those grounds are adequate. (Feldman and Coneehold this view.)

All of the preceding views count as internalist because they endorse Simple

Internalism. The following views reject Simple Internalism:

(iv) A weak form of externalism requires every justi®ed belief to be based

on some ground whose presence is accessible, but tells an externalist

story about what it is for grounds to be adequate. (This is the view

held by Alston and some other externalists; see fn. 21.)

108 James Pryor

21 That is Alston's view (see [1989], Essays 4 and 9). See also Goldman ([1980]); Swain ([1981], Ch.

4, [1985]); and Conee ([1988]). Such views are examples of what Pollock and Cruz call `norm

externalism' ([1999], Ch. 5).22 A further question about access is: Do we have special access to facts about whether a belief is based

on a given ground? For discussion of this question, see Audi ([1993], Ch. 11) and Kim ([1993]).



(v) The strongest form of externalism not only denies Simple Internal-

ism, but also denies that justi®ed beliefs are always based on

accessible grounds. (Plantinga advocates a view of this sort about

warrant; I will discuss the dierence between justi®cation and

Plantinga's notion of `warrant' in Section 4.2, below.)

Of these views, (i), (iv), and (v) have been most prominent in the literature. (ii)

and (iii) have had fewer champions. These latter two positions are conceptually

important, however, for if one overlooked them, one might take criticisms of the

other positions to have more force than they actually do. For instance,

arguments against (i) cannot by themselves compel one to accept (iv) or (v); nor

can arguments for Simple Internalism by themselves compel one to accept (i).

Now we understand the internalism/externalism debate a bit better, and

appreciate the variety of forms it can take. Let's turn next to the question

whether assumptions about the ethics of belief force us to take any particular

position in the internalism/externalism debate.

4 The ethics of belief

4.1 What does it mean to say that justi®cation is `normative'?

It is often argued that the fact that justi®cation is a certain kind of normative

notion has implications for the internalism/externalism debate. I want to

examine this claim. To begin, we need to know what's meant by saying that

justi®cation is `normative', or that epistemology is a `normative' discipline.

Here we face the problem that there are dierent ways of being `normative',

important to dierent epistemological debates. We need to separate out the

ones that are relevant here.23

The ®rst debate about `normativity' concerns the question of how relevant

empirical research is to epistemology. Some naturalists argue that epistemol-

ogy should be conceived as an extension of descriptive psychology; theiropponents argue that epistemology is more than merely a descriptive

enterprise, and that the proper methodology for doing epistemology is an a

priori one. Various intermediate views are possible.24


23 In constructing this chart of dierent debates about `normativity', I've been much helped by:

Alston ([1989], Essays 4, 5, and 8); Kornblith ([1983]); Feldman and Conee ([1985]); Feldman

([1988], [2000]); Goggans ([1991]); Plantinga ([1993b], pp. 45.); Montmarquet ([1993],

Appendix 2); Foley ([1994]); Goldman ([1994]); Fumerton ([1995], Ch. 1); and discussions with

J. Cruz, T. Kelly, and R. Wedgwood.24 Quine ([1969]) is standardly thought to have favored replacing epistemology with descriptive

psychology. (See Foley [1994] for a careful account of what Quine's actual views were.) Othernaturalists like Kornblith don't go so far; they believe there are genuine normative facts about

which modes of inference are good modes of reasoning; but they think we can't tell what those

good reasons are without studying how we actually do reason and acquire beliefs. For some

useful orientation to this debate, see Kim ([1988]); Mae ([1990b]); Goldman ([1994]); BonJour

([1994]); Kornblith ([1995], [1999]); and Feldman ([1999b]).



The second debate about `normativity' in epistemology concerns the

question whether epistemological recommendations are merely hypothetical ,

or whether they're somehow non-optional or categorical . Some philosophers

argue that it's only insofar as you have an independent interest (perhaps a

prudential interest) in having `epistemically rational' beliefs, or getting at the

truth in a reliable way, that you should care what beliefs are proscribed or

recommended by epistemology.25 Their opponents argue that epistemology's

recommendations apply to all agents, independently of what independent

interests those agents may have. (Of course, an agent can be unaware of some

epistemological prescription that does in fact apply to him.)

A third debate about `normativity' focuses around the kinds of recommenda-

tions that epistemology should be seen as issuing. To some extent, we can aect

what we believe by the actions we take.26 So the question arises: does epistemologymerely issue guidelines for evaluating the beliefs we end up with, or does it also

issue guidelines for how to go about forming beliefs? For instance, does it instruct

us not to perform actions that would result in our having Pascal's-wager-type

beliefs, beliefs which are bene®cial but evidentially unsupported?27 And does

epistemology only concern itself with what we ought to believe, given a certain

body of evidence? Or does it also issue guidelines on how to behave as an inquirer?

For instance, does it ever tell us how and when to go out and gather more

evidence? Does it tell us to consult multiple sources, use controls in our

experiments, refrain from destroying or manufacturing evidence, listen carefully

to criticism, avoid mind-altering drugs, and so on?28 And if epistemology does

issue guidelines of this sort, the question arises: in what ways does a subject's

conforming or failing to conform to those guidelines aect the justi®catory status

of his beliefs? (I place this last question under the heading of the ®fth debate,

discussed below.)

A fourth debate about `normativity' concerns the connection between being

justi®ed and following the norms or guidelines that epistemology issues. Is the

notion of justi®cation supposed to play a role in guiding and regulating ourbeliefs? That is, ought we to say that a belief is justi®ed when and only when the

110 James Pryor

25 Some relevant literature here is Laudan ([1987]); Foley ([1987]); Giere ([1988]); Stich ([1990]);

Mae ([1990a]); Kornblith ([1993]); and Nozick ([1993]).26 There is an extensive literature on what sorts of choice or voluntary control we have in forming

our beliefs. See e.g. Williams ([1970]); Winters ([1979]); Heil ([1983b], [1984]); Naylor ([1985]);

Cook ([1987]); Alston ([1989], Essay 5); Feldman ([1988], [2001]); Bennett ([1990]);

Montmarquet ([1993], Ch. 5); Plantinga ([1993a], Ch. 2); Scott-Kakures ([1994]); Radclie

([1997]); Steup ([1988], [1999]); and various essays in Steup ([2001]).27 This issue is connected to the question discussed above, whether epistemological recommenda-

tions have more than hypothetical force. See also Meiland ([1980]); Heil ([1983a], [1992]); Kelly([2000]); and Feldman ([2000]).

28 See Goldman ([1978], [1991]); Kornblith ([1983]); Heil ([1983b]); Feldman and Conee ([1985]);

Foley ([1987]); Laudan ([1996]); Hall and Johnson ([1998]); Feldman ([1988], [2000]); and

DeRose ([2000]) for discussion of these kinds of epistemic guidelines. Some of the literature on

virtue epistemology is also relevant here; see the surveys cited in fn. 2.



belief is permitted or recommended by some proper belief-guiding recipe, which

the agent adhered to in forming the belief? Or ought we to view the notion of

justi®cation as merely playing a role in third-person evaluations of belief, e.g.

when we're deciding what propositions some agent counts as knowing? It is

controversial whether the notion of justi®cation should be thought of as a belief-

guiding notion.29 One's answer to that question will depend in part on how much

choice or control one thinks we have over what we believe (see fn. 26). Some

philosophers have argued that the question whether justi®cation is a belief-

guiding notion profoundly aects the internalism/externalism debate. We will

consider the extent to which it does aect that debate below. But we need a bit

more conceptual apparatus before we can do that. So let's press on.

The ®fth and last debate about `normativity' in epistemology concerns the

connection between justi®cation and the notions of epistemic praise andblame. Some philosophers believe that justi®cation should be understood in

terms of epistemic blamelessness and epistemic responsibility. They say that

whether one is justi®ed depends on how well one has met one's epistemic

obligations, whether one is `doing the best one can' as a believer, or at least

the best that can reasonably be expected of one, and so on. In the literature,

these are called deontological conceptions of justi®cation, because of the

primacy they usually attach to meeting one's epistemic obligations. We might

also speak of praise-and-blame conceptions of justi®cation, or responsibilist

conceptions of justi®cation, if we wanted to emphasize dierent elements.

These views are all closely related to each other.30

Philosophers who oppose those views acknowledge that `justi®cation' is an

evaluative notion, a term of praise or criticism; but they deny that there is a tight

connection between one's justi®catory status and how well one has met one's

obligations, whether one merits praise and blame, and so on. Some of these

opponents deny that we have any epistemic obligations. Others agree that we have

epistemic obligationsÐand that we can assign a kind of epistemic praise and


29 Goldman is an example of someone who construes justi®cation as a merely evaluative notion,

not a belief-guiding one. (Goldman does sometimes discuss regulative questions, though: see

[1978], [1980]. Note that the `rules' he talks about in [1986] are not meant to constitute a belief-

guiding recipe. Goldman emphasizes this on pp. 25±26 and 59.)30 For discussions of the conceptual connections between justi®cation, epistemic obligation, and

epistemic praise and blame, see: Goldman ([1980], esp. Section V, [1988]); Plantinga ([1988],

[1993a], Ch. 1); Steup ([1988]); Alston ([1989], Essays 4, 5, and 8); Moser ([1989], Section 1.3);

Fumerton ([1995], Ch. 1); and Haack ([1997]). Some philosophers who think of justi®cation in

deontological terms are: Ginet ([1975], Ch. 3); Chisholm ([1977], Ch. 1, Section 5, [1982]);

Kornblith ([1983]); BonJour ([1985], Ch. 1); and Steup ([1999]). See the start of Goldman

([1999]) on the connection between deontologism and the belief-guiding conception of

justi®cation described above.In discussions of deontological conceptions of justi®cation, it's common to distinguish

between objective obligations and subjective obligations, and to tie blamelessness to conformity

to one's subjective obligations. See Goldman ([1980], pp. 36., [1986], pp. 73.); Alston ([1989],

Essay 4 pp. 86.); Plantinga ([1993a], pp. 15.); and Pollock and Cruz ([1999], pp. 140.) for

this distinction.



blame by reference to themÐbut they deny that one's justi®catory status is always

a function of the epistemic praise or blame one merits. On such views, the

question whether one's belief is genuinely justi®ed is not settled by the fact that

one is blameless in holding the belief. As they see it, there is a one kind of epistemic

standing, which involves meeting all of one's epistemic obligations and

proceeding in a responsible and blame-free manner; and there is another kind

of epistemic standing, having a genuinely justi®ed belief; and the ®rst does not

suce for the second. (We will discuss such views further below.)

In what follows, we will focus on whether justi®cation is `normative' in the

ways operative in the fourth and ®fth debates, and how one's views about

that aect the internalism/externalism debate.

4.2 Justi®cation, warrant, and epistemic blamelessness

These days, unfortunately, the terminological situation with respect to

`justi®cation' is rather messy. A bit of stipulation will help.

If you believe that Bing is a male canary, and that 9 out of 10 male canaries

can sing, and you have no other relevant evidence, then regardless of what

you're inclined to believe on the basis of that evidence, there's an intuitive

sense in which your evidence supports the conclusion that Bing can sing.

Similarly, if you have a visual experience as of something yellow, and no

countervailing evidence, what your evidence supports is the belief that there issomething yellow in front of you. Let's call a belief that is supported in this

way a justi®catorily supported belief. For now, let's leave it an open question

whether this support relation is internalist or not; and let's also leave it open

whether this support relation should be understood in deontological terms.

I ®nd it most natural to use the term `justi®cation' to refer to the support

relation just articulated, however that relation ultimately gets to be analyzed.31

However, as we'll see, some philosophers use the term `justi®cation' in a more

restrictive way.Following Plantinga, we can reserve the term `warrant' for a second

notion, the notion of what has to be added to true belief to yield knowledge.32

112 James Pryor

31 Alston ([1993]) and Cohen ([1995]), on the other hand, doubt whether there is any common pre-

theoretic notion of justi®catory support for internalists, externalists, deontologists, etc., to be

giving competing analyses of.32 See Plantinga ([1993a], [1993b]). Other philosophers use the term `warranted' to express dierent

notions (see e.g. Meiland [1980]; Pollock [1983]; and Wright [1991]).

Burge ([1993], [1996]) draws a contrast between `entitlement' and `justi®cation'. In Burge's

usage, justi®cation is a kind of positive epistemic quality whose presence and nature is readily

accessible to the subject upon re¯ection. Entitlement is a lower-grade positive epistemic quality

whose presence and nature is not as readily accessible. On Burge's view, basic unre ectivecapacities like perception give us entitlements for believing things, but not justi®cations. Burge

uses the term `warrant' to describe the entire class of positive epistemic qualities, comprising both

entitlements and justi®cations. In my terminology, what Burge calls `warrants' are relations of

justi®catory support, and what he calls `justi®cations' and `entitlements' are two dierent species

of justi®catory support. Burge does not discuss the Plantingean notion of warrant at all.



It is most natural to think that warranted beliefs are beliefs for which one has

justi®catory support, which are appropriately based, and which have some

extra features necessary to handle the Gettier problem. But there is much

disagreement about this. Some philosophers deny that warrant can be

analyzed into justi®catory support plus some extra features.33 And some

philosophers believe that it's possible to know things for which one has no

justi®catory support (see e.g. Alston [1989], Essay 7); if they're right, it

follows that one can sometimes have warranted beliefs for which one has no

justi®catory support.

As I said above, not every philosopher acknowledges the existence of

epistemic obligations. But as I also pointed out, even if one does acknowledge

the existence of such obligations, it's a further question what the relationship is

between meeting one's epistemic obligations and having a justi®catorilysupported belief. So let's stipulate that talk of `epistemic blamelessness' is to

concern how well one meets one's epistemic obligations, behaves in an

epistemically responsible manner, and so on. Our stipulation will leave it open

whether being justi®ed and being epistemically blameless always go hand-in-

hand. It needs argument to show that there is any close connection between

these properties. Deontologists think such an argument can be given; and hence

that being epistemically blameless suces for being justi®ed. We'll look at some

reasons for disagreeing in a moment. But however that debate turns out, we can

all agree that there is a legitimate and useful notion of epistemic blamelessness.

As I said, some philosophers deny that we have any real epistemic obligations.

Hence they will not think that the notions of epistemic praise and blame I'm

describing have any application. They might hold this view because they think we

lack the kind of voluntary control over our beliefs that the notions of praise and

blame demand (see fn. 26, above). My point is just that one can think there is a

legitimate and useful notion of epistemic blamelessness without thereby

committing oneself to a deontological conception of justi®cation.

For example, Montmarquet describes a notion of `justi®cation' that hasclose ties to epistemic praise and blame, but he wants this notion of

justi®cation to ground our ascriptions of moral praise and blame for actions,

not to play any important role in the theory of knowledge (see [1993],

esp. Ch. 6). In my terminology, Montmarquet is giving an account of

epistemic blamelessness, and not taking a stand on what the relation is

between justi®cation and blamelessness.

Later in his career, Chisholm also advocated the view that we have

epistemic obligations, but the notion of justi®cation should not be explicated

in terms of them.34


33 This is a guiding idea in Williamson ([1995], [1997]) and Plantinga ([1993a], [1993b]).34 Plantinga calls this `post-classical Chisholm' (see Plantinga [1993a], Ch. 3).



Plantinga reserves the term `justi®cation' for the notion I'm calling

`epistemic blamelessness'; and Alston is tempted to do the same.35 They

claim that the etymology of `justi®cation' supports understanding it in

this deontological way. I think it's unclear what the etymology supports;

but in any case, in natural languages usage tends not to be a slave to

etymology. And given the current usage of the term `justi®cation' among

epistemologists, it seems to me better to use the term `justi®cation' for

our ®rst notion, the relation of justi®catory support whose nature is

being debated. Many past and some present-day epistemologists believe

that justi®cation should be explicated in deontological terms. But we

shouldn't build it into our de®nition of what we mean by `justi®cation'

that this is so.

4.3 How should one's ethics of belief aect one's views about the

internalism/externalism debate?

Now we are ®nally in a position to explore the connections between the ethics

of belief and the internalism/externalism debate. These connections have been

the subject of much recent discussion.

There are good reasons for thinking that justi®cation and epistemic

blamelessness can come apart. Doing all that can be reasonably be

expected of you, epistemically, is compatible with holding beliefs that have

little or no genuine support. Imagine a hapless subject who is taught bad

epistemic standards. For instance, suppose he makes mistakes when

engaging in tricky statistical reasoning. Let's say the standards he's been

taught fail to distinguish between the likelihood that a test will yield a

false negative and the likelihood that the test will yield a false positive.Our hapless subject does his best to apply these standards, but

unfortunately he is not intellectually capable of discerning their defects.

Furthermore, he has lost all memory of his childhood and so doesn't recall

how it was he ®rst acquired these standards. Cases of this sort naturally

prompt two judgments. First, the subject's beliefs are epistemically

defective in some important way. But second, the subject cannot be held

culpable for those defects. You can say `The hapless subject ought to

believe the things that his evidence really supports', but this `ought'

doesn't seem to express any epistemic obligation. Our subject is doing the

best he can, and the best that can reasonably be demanded of him. That

114 James Pryor

35 See Plantinga ([1993a]) and Alston ([1989], Essay 4, fn. 21 and Essay 5, p. 143). Goldman

([1988]) calls the notion of epistemic blamelessness `weak justi®cation'.



seems a good reason to say he's violating no epistemic obligations.36 Yet

the beliefs he forms on the basis of statistical reasoning will be seriously

defective; and for that reason it's natural to regard those beliefs as

unjusti®ed . Hence it appears that a subject can sometimes be blameless for

holding unjusti®ed beliefs.37

I myself think that justi®cation and epistemic blamelessness do come apart in

this way, and that this militates against any deontological conception of

justi®cation. But for the sake of argument, let's suppose that the line of

reasoning I sketched can be resisted, and that the deontologist is right to forge a

strong connection between justi®cation and epistemic blamelessness. What then

would follow? One often sees it claimed that deontologism and internalism go

hand-in-hand. However, there are some dissenting voices about this.

Let's ®rst consider whether viewing justi®cation as a kind of belief-guidingnotion, or viewing it in deontological terms, commits one to internalism.38


36 Feldman (Feldman and Conee [1985]; Feldman [1988], [2000]) and Wolterstor ([1997]) employ

a very thin notion of òbligation', which merely requires some epistemic òught' -claim to be

true of a subject. This thin notion is insensitive to issues about what can reasonably be

demanded of a subject, and whether the subject could have done otherwise than he did. I think

this notion is too thin to capture our concept of a genuine obligation. (See Goggan's [1991]

useful contrast between `purely valuative' norms and `deontic' norms. Only the second kind of

norm corresponds to a genuine kind of epistemic obligation.) In any case, regardless of whether

we regard this thin notion as genuine kind of obligation, it should be clear that there is no directroute from what one is òbliged' to do in that sense to what praise and blame one merits. When

we talk of epistemic praise and blame, I assume we're talking about a kind of genuine

culpability on the part of the subject, and not merely complimenting or criticizing the subject's

beliefs. Praise and blame in this sense are sensitive to what can reasonably be demanded of one,

and whether one could have done otherwise. In ([2000]), Feldman acknowledges that the notion

of òbligation' he's employing may come apart from assessments of praise and blame.37 Alston ([1989], Essay 4); Goldman ([1988]); and Feldman and Conee ([forthcoming]) all defend

the claim that being epistemically blameless does not suce for being justi®ed. (Goldman calls

blameless beliefs `weakly justi®ed', and calls genuinely justi®ed beliefs `strongly justi®ed'.)

Alston and Goldman believe that the notion of being genuinely justi®ed, which goes beyond

mere blamelessness, should be analyzed in externalist terms. However, I take that to be a

further question. For a nice critical summary of Goldman's views on this topic, see Riggs

([1997]).

Plantinga ([1988], [1993a], Chs 1±2) argues that epistemic blamelessness doesn't suce for

warranted belief; but many of the cases he discusses also show that being epistemically

blameless doesn't suce for having what I'm calling a genuinely justi®ed belief. (Be careful:

Plantinga uses the term `justi®ed' dierently than me; he uses it as a synonym for èpistemically

blameless'.)

In the text, I presented an argument that one can have unjusti®ed beliefs that are

epistemically blameless. Another question is whether it's possible to have blameworthy beliefs

that are none the less justi®ed. That is a more delicate question.38 Plantinga ([1993a], Ch. 1) argues that it would commit one to internalism. Alston ([1989], Essay

8) discusses arguments from deontologism to dierent internalist theses. See also Greco

([1990]). Pollock and Cruz ([1999], Ch. 5) argue from a regulative or `procedural' conception of

justi®cation to a kind of ìnternalism'. However, their de®nition of ìnternal state' (see pp. 132± 33) is dierent from the more standard one being used here; as a result, we do not mean the

same thing by ìnternalism'. Also, I'm not convinced that their understanding of what it is for

behavior to be `regulated by a norm' corresponds to the subject's genuinely following, or being

guided by a norm. They seem to be more concerned with the underlying mechanisms that

causally control a subject's behavior.



Let's assume that the relevant belief-guiding recipes or norms or obligations

are of the form: `In circumstances C, believe p'. The question then arises

whether C has to be the kind of circumstance to which the subject has the

`special access' that is operative in the internalism/externalism debate. For

instance, must the subject always be able to tell whether C obtains by

re¯ection alone? Deontologists often argue in the following way:

If a belief-guiding recipe is to be usable in deciding what to believe, then

the circumstances C it refers to must be circumstances such that the

subject can tell whether they obtain, when he's following the recipe. In

particular, they must be such that the subject can tell by re¯ection alone

whether they obtain.

However, Goldman argues that this last claim is gratuitous. Perhaps the

subject needs to be able to tell whether C obtains, if he's to follow the recipe,

but Goldman denies that he has to be able to tell whether C obtains by

re¯ection alone (see Goldman [1980], esp. Section IX, [1999]; Jacobson

[1997]). So the extent to which we have `special access' to C is a debated

question.

Another question that arises is whether the fact that a given belief-guiding

recipe is correct will be an internally determined matter, or a matter that one

can ascertain by re¯ection alone. These issues are also controversial. Forinstance, Goldman at one point defended a reliabilist story about what makes

belief-guiding recipes correct (again, see Goldman [1980] and Jacobson

[1997]).

Finally, Kornblith ([1982]) argues that thinking that justi®cation is to be

explained in terms of epistemological responsibility does not commit us to

any form of Access Internalism.

So there does not seem to me to be any straightforward, uncontroversial

route from a deontological or belief-guiding conception of justi®cation to any

of the internalist views we described earlier.

Let's instead consider whether there are any interesting entailments in the

opposite direction. Plantinga argues that internalism is well motivated only

when one is working with a deontological conception of justi®cation ([1993a],

esp. Chs 1 and 3). Is that right?

It is true that some prominent rationales for internalism do employ

deontological conceptions of justi®cation. For instance, arguments from

thought-experiments involving brains in vats often include remarks like this:

There's no way for the brain to discover his plight; so he can't be blamed

for forming the beliefs he does. He forms the same beliefs anyone else

would form on the basis of that evidence. So even though many of his

beliefs are false, he is at least justi®ed in having those beliefs.

116 James Pryor



Here it's being assumed that being epistemic blameless is sucient for being

justi®ed.

But while arguments for internalism often do employ deontological

conceptions of justi®cation, I am not persuaded that they have to. For

instance, Cohen ([1984]) argues for internalism by appeal to thought

experiments involving brains in vats, and his arguments never appeal in

any essential way to deontological assumptions. Many philosophers share

Cohen's intuition that it's possible for a brain in a vat, if he conducts his

aairs properly, to have many justi®ed (albeit false) beliefs about his

environment. AndÐat least in my caseÐthis intuition survives the

recognition that being epistemically blameless does not suce for being

justi®ed. It doesn't merely seem to be the case that the brain in a vat can form

beliefs in a way that is epistemically blameless. It also seems to be the casethat he can form beliefs in a way that is epistemically proper, and that the

beliefs he so forms would be fully justi®edÐdespite the fact that they're

reliably false.

To illustrate this, consider three brains in a vat. The ®rst brain is

epistemically quite reckless. He believes whatever he feels like believing, so

long as he doesn't have incontrovertible evidence that it's false. The second

brain is like the hapless subject we considered above. He's doing the best he

can, epistemically, and the best that can be reasonably expected of him. He's

very scrupulous and careful in following the epistemic standards he picked up

as a child. Unfortunately, these epistemic standards are defective in a way

he's not capable of recognizing. (Perhaps, as before, they fail to distinguish

false negatives from false positives.) The third brain carefully follows the

same epistemic standards we follow.

It seems natural to say that, although there's a sense in which the second

and third brains are both epistemically blameless, only the third is conducting

his epistemic aairs properly, and so only he is justi®ed in his beliefs. This

indicates that there's a kind of positive epistemic status which goes beyondmere epistemic blamelessness, but which one can possess even if one's beliefs

are reliably false, as the brains' beliefs are. Such considerations incline me to

an internalist, non-deontological account of justi®cation.39

Now, not every philosopher shares these intuitions about brains in vats. So

by themselves they are not likely to settle the debate about internalism. Other

arguments are needed to do that. But these intuitions about brains in vats do

cast doubt on the claim that it would only be possible to motivate or defend

internalism by appealing to deontological assumptions. I have so far seen no


39 This style of argument from multiple brains in vats derives from Cohen ([1984], p. 283). There

are a number of other non-deontological internalists. See e.g. Fumerton ([1988], [1995], Ch. 1);

Audi ([1993], Ch. 10); and Feldman and Conee ([forthcoming]). See also the works of `post-

classical Chisholm', referred to in fn. 33, above.



convincing argument for that claim. So there do not seem to me to be any clear

entailments from internalism to a deontological conception of justi®cation.

In my own view, then, it remains to be seen whether there is any tight

connection between the ethics of belief and the internalism/externalism

debate. I certainly have not shown that there could be no such connection.

But I hope that the present discussion and the papers I've pointed to make it

clear that this question is really less settled than it sometimes appears.

Acknowledgements

I would like to thank J. Cruz, T. Kelly, and R. Feldman for their extremely

helpful advice and feedback.

Department of Philosophy

Harvard University

Cambridge, MA 02138, USA

[email protected]

References

Alston, W. [1989]: Epistemic Justi®cation, Ithaca: Cornell University Press.

Alston, W. [1993]: Èpistemic Desiderata', Philosophy and Phenomenological Research,

53, pp. 527±51.

Annis, D. [1978]: À Contextualist Theory of Epistemic Justi®cation', American

Philosophical Quarterly, 15, pp. 213±19.

Audi, R. [1993]: The Structure of Justi®cation, Cambridge: Cambridge University

Press.

Axtell, G. [1997]: `Recent Work on Virtue Epistemology', American Philosophical

Quarterly, 34, pp. 1±26.

Bealer, G. [1987]: `Philosophical Limits of Scienti®c Essentialism', in J. Tomberlin

(ed.), Philosophical Perspectives, 1, pp. 289±365.Bealer, G. [1999]: `The A Priori', in J. Greco and E. Sosa ( eds), Blackwell Guide to

Epistemology, Malden, MA: Blackwell, pp. 243±70.

Bennett, J. [1990]: `Why Is Belief Involuntary?' Analysis, 50, pp. 87±107.

Boghossian, P. and Peacocke, C. [2001]: New Essays on the A Priori , Oxford: Oxford

University Press.

Bonjour, L. [1978]: `Can Empirical Knowledge Have a Foundation?', American


Bonjour, L. [1985]: The Structure of Empirical Knowledge, Cambridge, MA: Harvard

University Press.Bonjour, L. [1994]: Àgainst Naturalized Epistemology', in P. A. French, T. E.

Uehling and H. Wettstein (eds), Midwest Studies in Philosophy, 19, pp. 283±300.

Bonjour, L. [1998]: In Defense of Pure Reason, Cambridge: Cambridge University

Press.

118 James Pryor



Bonjour, L. [1999a]: `The Dialectic of Foundationalism and Coherentism', in J. Greco

and E. Sosa (eds), The Blackwell Guide to Epistemology, Malden, MA: Blackwell,

pp. 117±42.

Bonjour, L. [1999b]: `Foundationalism and the External World', in J. Tomberlin (ed.),

Philosophical Perspectives, 13, pp. 229±50.

Burge, T. [1993]: `Content Preservation', Philosophical Review, 102, pp. 457±88.

Burge, T. [1996]: Òur Entitlement to Self-Knowledge', Proceedings of the Aristotelian

Society, 96, pp. 91±116.

Casullo, A. [1988]: `Revisability, Reliabilism and A Priori Knowledge', Philosophy and

Phenomenological Research, 49, pp. 187±213.

Casullo, A. [1999]: A Priori Knowledge, Aldershot: Ashgate/Dartmouth.

Chisholm, R. M. [1977]: Theory of Knowledge, 2nd edn, Englewood Clis, NJ:

Prentice-Hall.

Chisholm, R. M. [1982]: À Version of Foundationalism', in Foundations of Knowing,

Minneapolis: University of Minnesota Press, pp. 3±32.

Cohen, S. [1984]: `Justi®cation and Truth', Philosophical Studies, 46, pp. 279±95.

Cohen, S. [1987]: `Knowledge, Context, and Social Standards', Synthese, 73, pp. 3±26.

Cohen, S. [1988]: `How to Be a Fallibilist', in J. Tomberlin (ed.), Philosophical

Perspectives, 2, pp. 91±123.

Cohen, S. [1991]: `Skepticism, Relevance, and Relativity', in B. McLaughlin (ed.),

Dretske and His Critics, Cambridge, MA: Blackwell, pp. 17±37.

Cohen, S. [1995]: Ìs There an Issue About Justi®ed Belief?', Philosophical Topics, 23,pp. 113±27.

Cohen, S. [1998]: `Fumerton on Metaepistemology and Skepticism', Philosophy and


Cohen, S. [1999]: `Contextualism, Skepticism, and the Structure of Reasons', in J.

Tomberlin (ed.), Philosophical Perspectives, 13, pp. 57±90.

Conee, E. [1988]: `The Basic Nature of Epistemic Justi®cation', Monist, 71, pp. 389±

404.

Cook, J. T. [1987]: `Deciding to Believe without Self-Deception', Journal of

Philosophy, 84, pp. 441±46.

Davidson, D. [1981]: À Coherence Theory of Truth and Knowledge', in E. LePore

(ed.) [1986], Truth and Interpretation: Perspectives on the Philosophy of Donald

Davidson, Oxford: Blackwell, pp. 307±19.

Davies, M. [1998]: Èxternalism, Architecturalism, and Epistemic Warrant', in C.

Wright, M. Smith and C. Macdonald (eds), Knowing Our Own Minds: Essays in

Self-Knowledge, Oxford: Oxford University Press, pp. 321±61.

Davies, M. [2001]: Èxternalism and Armchair Knowledge', in P. Boghossian and C.

Peacocke (eds), New Essays on the A Priori , Oxford: Oxford University Press.DeRose, K. [1992]: `Contextualism and Knowledge Attributions', Philosophy and


DeRose, K. [1995]: `Solving the Skeptical Problem', Philosophical Review, 104, pp. 1±

52.




DeRose, K. [1999]: `Contextualism: An Explanation and Defense', in J. Greco and E.

Sosa (eds), The Blackwell Guide to Epistemology, Malden, MA: Blackwell, pp. 187±

205.

DeRose, K. [2000]: Òught We to Follow Our Evidence?' Philosophy and


Dretske, F. [1970]: Èpistemic Operators', Journal of Philosophy, 67, pp. 1007±23.

Dretske, F. [1981]: `The Pragmatic Dimensions of Knowledge', Philosophical Studies,

40, pp. 363±78.

Dretske, F. [1991]: `Dretske's Replies', in B. McLaughlin (ed.), Dretske and His Critics,

Cambridge, MA: Blackwell.

Edidin, A. [1984]: À Priori Knowledge for Fallibilists', Philosophical Studies, 46,

pp. 189±97.

Fales, E. [1996]: A Defense of the Given, Lanham, MD: Rowman & Little®eld.

Feldman, R. [1988]: Èpistemic Obligations', in J. Tomberlin (ed.), Philosophical Perspectives, 2, pp. 235±56.

Feldman, R. [1999a]: `Contextualism and Skepticism', in J. Tomberlin (ed.),


Feldman, R. [1999b]: `Methodological Naturalism in Epistemology', in J. Greco and

E. Sosa (eds), The Blackwell Guide to Epistemology, Malden, MA: Blackwell,

pp. 170±86.

Feldman, R. [2000]: `The Ethics of Belief', Philosophy and Phenomenological Research,

60, pp. 667±96.

Feldman, R. [2001]: `Voluntary Belief and Epistemic Evaluation', in M. Steup (ed.),Knowledge, Truth and Duty: Essays on Epistemic Justi®cation, Responsibility, and

Virtue, Oxford: Oxford University Press.

Feldman, R. and Conee, E. [1985]: Èvidentialism', Philosophical Studies, 48, pp. 15±

34.

Feldman, R. and Conee E. [forthcoming]: Ìnternalism Defended', in H. Kornblith

(ed.), Epistemology: Internalism and Externalism, URL: <http://

www.ling.rochester.edu/feldman/papers.html>.

Foley, R. [1987]: The Theory of Epistemic Rationality, Cambridge, MA: Harvard

University Press.Foley, R. [1994]: `Quine and Naturalized Epistemology', in P. A. French, T. E.

Uehling and H. Wettstein (eds), Midwest Studies in Philosophy, 19, pp. 243±60.

Forbes, G. [1984]: `Nozick on Skepticism', Philosophical Quarterly, 34, pp. 43±52.

Fuller, S. [1996]: `Recent Work on Social Epistemology', American Philosophical


Fumerton, R. [1985]: Metaphysical and Epistemological Problems of Perception,

Lincoln, NB: University of Nebraska Press.

Fumerton, R. [1988]: `The Internalism/Externalism Controversy', in J. Tomberlin

(ed.), Philosophical Perspectives, 2, pp. 443±59.Fumerton, R. [1995]: Metaepistemology and Skepticism, Lanham, MD: Rowman &

Little®eld.

Giere, R. [1988]: Explaining Science, Chicago: University of Chicago Press.

Ginet, C. [1975]: Knowledge, Perception, and Memory, Dordrecht: Reidel.

120 James Pryor



Goggans, P. [1991]: Èpistemic Obligations and Doxastic Voluntarism', Analysis, 51,

pp. 102±5.

Goldman, A. [1976]: `Discrimination and Perceptual Knowledge', Journal of


Goldman, A. [1978]: Èpistemics: The Regulative Theory of Cognition', Journal of


Goldman, A. [1979]: `What Is Justi®ed Belief?' in G. Pappas (ed.), Justi®cation and

Knowledge, Dordrecht: Reidel, pp. 1±23.

Goldman, A. [1980]: `The Internalist Conception of Justi®cation', in P. A. French, T.

E. Uehling and H. Wettstein (eds), Midwest Studies in Philosophy, 5, pp. 27±51.

Goldman, A. [1986]: Epistemology and Cognition, Cambridge, MA: Harvard

University Press.

Goldman, A. [1988]: `Strong and Weak Justi®cation', in J. Tomberlin (ed.),


Goldman, A. [1991]: Èpistemic Folkways and Scienti®c Epistemology', in Liaisons,

Cambridge, MA: MIT Press, pp. 155±75.

Goldman, A. [1994]: `Naturalistic Epistemology and Reliabilism', in P. A. French, T.

E. Uehling and H. Wettstein (eds), Midwest Studies in Philosophy, 19, pp. 301±20.

Goldman, A. [1999]: Ìnternalism Exposed', Journal of Philosophy, 96, pp. 271±93.

Greco, J. [1990]: Ìnternalism and Epistemically Responsible Belief', Synthese, 85,

pp. 245±77.

Greco, J. [1993]: `Virtues and Vices of Virtue Epistemology', Canadian Journal of


Greco, J. [1999]: `Virtue Epistemology', Stanford Encyclopedia of Philosophy, URL:

<http://plato.stanford.edu/entries/epistemology-virtue/>.

Haack, S. [1993]: Evidence and Inquiry, Oxford: Blackwell.

Haack, S. [1997]: ```The Ethics of Belief'' Reconsidered', in L. E. Hahn (ed.), The

Philosophy of Roderick Chisholm, La Salle, IL: Open Court, pp. 129±44.

Hall, R. J. and Johnson, C. R. [1998]: `The Epistemic Duty to Seek More Evidence',

American Philosophical Quarterly, 35, pp. 129±39.

Harman, G. [1968]: `Knowledge, Inference, and Explanation', American Philosophical


Harman, G. [1973]: Thought, Princeton: Princeton University Press.

Harman, G. [1980]: `Reasoning and Evidence One Does Not Possess', in P. A. French,

T. E. Uehling and H. Wettstein (eds), Midwest Studies in Philosophy, 5, pp. 163±82.

Heil, J. [1983a]: `Believing What One Ought', Journal of Philosophy, 80, pp. 752±65.

Heil, J. [1983b]: `Doxastic Agency', Philosophical Studies, 43, pp. 355±64.

Heil, J. [1984]: `Doxastic Incontinence', Mind , 93, pp. 56±70.

Heil, J. [1992]: `Believing Reasonably', NouÃs, 26, pp. 47±62.

Heller, M. [1989]: `Relevant Alternatives', Philosophical Studies, 55, pp. 23±40.Heller, M. [1999a]: `The Proper Role for Contextualism in an Anti-Luck

Epistemology', in J. Tomberlin (ed.), Philosophical Perspectives, 13, pp. 115±30.

Heller, M. [1999b]: `Relevant Alternatives and Closure', Australasian Journal of





Jacobson, S. [1997]: Èxternalism and Action-Guiding Epistemic Norms', Synthese,

110, pp. 381±97.

Kelly, T. [2000]: `The Ethics of Belief', Ph.D. Thesis, Harvard University.

Kim, J. [1988]: `What Is ``Naturalized Epistemology``?' in J. Tomberlin (ed.),


Kim, K. [1993]: Ìnternalism and Externalism in Epistemology', American


Kitcher, P. [1992]: `The Naturalists Return', Philosophical Review, 101, pp. 53±114.

Korcz, K. A. [1997]: `Recent Work on the Basing Relation', American Philosophical


Kornblith, H. [1982]: `The Psychological Turn', Australasian Journal of Philosophy, 60,

pp. 238±53.

Kornblith, H. [1983]: `Justi®ed Belief and Epistemically Responsible Action',

Philosophical Review, 92, pp. 33±48.Kornblith, H. [1993]: Èpistemic Normativity', Synthese, 94, pp. 357±76.

Kornblith, H. [1995]: `Naturalistic Epistemology and Its Critics', Philosophical Topics,

23, pp. 237±55.

Kornblith, H. [1999]: Ìn Defense of a Naturalized Epistemology', in J. Greco and E.

Sosa (eds), The Blackwell Guide to Epistemology, Malden, MA: Blackwell, pp. 158±

69.

Laudan, L. [1987]: `Progress or Rationality? The Prospects for Normative

Naturalism', American Philosophical Quarterly, 24, pp. 19±31.

Laudan, L. [1996]: Beyond Positivism and Relativism, Boulder, CO: Westview.Lewis, D. [1996]: Èlusive Knowledge', Australasian Journal of Philosophy, 74,

pp. 549±67.

Ludlow, P. and N. Martin. [1998]: Externalism and Self-Knowledge, Stanford: CSLI

Publications.

Luper-Foy, S. [1987]: The Possibility of Knowledge: Nozick and His Critics, Totowa,

NJ: Rowman & Little®eld.

McGinn, C. [1984]: `The Concept of Knowledge', in P. A. French, T. E. Uehling and

H. Wettstein (eds), Midwest Studies in Philosophy, 9, pp. 529±54.

McGrew, T. [1995]: The Foundations of Knowledge, Lanham, MD: Little®eld Adams.Mae, J. [1990a]: `Naturalism and the Normativity of Epistemology', Philosophical

Studies, 59, pp. 333±49.

Mae, J. [1990b]: `Recent Work on Naturalized Epistemology', American


Meiland, J. [1980]: `What Ought We to Believe? The Ethics of Belief Reconsidered',

American Philosophical Quarterly, 17, pp. 15±24.

Montmarquet, J. [1993]: Epistemic Virtue and Doxastic Responsibility, Lanham, MD:

Rowman & Little®eld.

Moser, P.K. [1985]: Empirical Justi®cation, Dordrecht: Reidel.Moser, P.K. [1987]: A Priori Knowledge, Oxford: Oxford University Press.

Moser, P.K. [1989]: Knowledge and Evidence, New York: Cambridge University Press.

Naylor, M.B. [1985]: `Voluntary Belief', Philosophy and Phenomenological Research,

45, pp. 427±36.

122 James Pryor



Nozick, R. [1981]: Philosophical Explanations, Cambridge, MA: Harvard University

Press.

Nozick, R. [1993]: The Nature of Rationality, Princeton: Princeton University Press.

Plantinga, A. [1988]: `Positive Epistemic Status and Proper Function', in J. Tomberlin

(ed.), Philosophical Perspectives, 2, pp. 1±50.

Plantinga, A. [1993a]: Warrant: The Current Debate, New York: Oxford University

Press.

Plantinga, A. [1993b]: Warrant and Proper Function, New York: Oxford University

Press.

Pollock, J. [1983]: Èpistemology and Probability', Synthese, 55, pp. 231±52.

Pollock, J. and Cruz, J. [1999]: Contemporary Theories of Knowledge, 2nd edn,

Totowa, NJ: Rowman & Little®eld.

Pryor, J. [2000]: `The Skeptic and the Dogmatist', NouÃs, 34, pp. 517±49.

Quine, W. V. O. [1969]: Èpistemology Naturalized', in Ontological Relativity and

Other Essays, New York: Columbia University Press, pp. 69±90.

Radclie, D. [1997]: `Scott-Kakures on Believing at Will', Philosophy and


Riggs, W. D. [1997]: `The Weakness of Strong Justi®cation', Australasian Journal of


Schier, S. [1996]: `Contextualist Solutions to Skepticism', Proceedings of the

Aristotelian Society, 96, pp. 317±33.

Schmitt, F. [1999]: `Social Epistemology', in J. Greco and E. Sosa (eds), The Blackwell Guide to Epistemology, Malden, MA: Blackwell, pp. 354±82.

Scott-Kakures, D. [1994]: Òn Belief and Captivity of the Will', Philosophy and


Sellars, W. [1956]: Èmpiricism and the Philosophy of Mind', in H. Feigl and M.

Scriven (eds), Minnesota Studies in the Philosophy of Science, 1, pp. 253±329.

Sosa, E. [1991]: Knowledge in Perspective, Cambridge: Cambridge University Press.

Steup, M. [1988]: `The Deontic Conception of Epistemic Justi®cation', Philosophical

Studies, 53, pp. 65±84.

Steup, M. [1999]: À Defense of Internalism', in L. Pojman (ed.), Theory of Knowledge:

Classical and Contemporary Readings, Belmont, CA: Wadsworth, pp. 373±84.

Steup, M. [2001]: Knowledge, Truth and Duty: Essays on Epistemic Justi®cation,

Responsibility, and Virtue, Oxford: Oxford University Press.

Stich, S. [1990]: The Fragmentation of Reason, Cambridge, MA: MIT Press.

Stine, G. C. [1976]: `Skepticism, Relevant Alternatives, and Deductive Closure',

Philosophical Studies, 29, pp. 249±61.

Swain, M. [1981]: Reasons and Knowledge, Ithaca: Cornell University Press.

Swain, M. [1985]: `Justi®cation, Reasons and Reliability', Synthese, 64, pp. 69±92.Triplett, T. [1990]: `Recent Work on Foundationalism', American Philosophical


Unger, P. [1986]: `The Cone Model of Knowledge', Philosophical Topics, 14, pp. 125±

78.




Vogel, J. [1990]: Àre There Counter-Examples to the Closure Principle?' in M. D.

Roth and G. Ross (eds), Doubting: Contemporary Perspectives on Skepticism,

Dordrecht: Kluwer, pp. 13±27.

Vogel, J. [1999]: `The New Relevant Alternatives Theory', in J. Tomberlin (ed.),

Philosophical Perspectives, 13, pp. 155±80.Williams, B. [1970]: `Deciding to Believe', in B. Williams [1973], Problems of the Self ,

Cambridge: Cambridge University Press, pp. 136±51.

Williams, M. [1996]: Unnatural Doubts, Princeton: Princeton University Press.

Williams, M. [1999a]: Groundless Belief , 2nd edn, Oxford: Blackwell.

Williams, M. [1999b]: `Skepticism', in J. Greco and E. Sosa (eds), The Blackwell Guide

to Epistemology, Malden, MA: Blackwell, pp. 35±69.

Williamson, T. [1995]: Ìs Knowing a State of Mind?', Mind , 104, pp. 533±65.

Williamson, T. [1997]: `Knowledge as Evidence', Mind , 106, pp. 717±41.

Winters, B. [1979]: `Believing at Will', Journal of Philosophy, 76, pp. 243±56.Wolterstor, N. [1997]: Òbligations of Belief: Two Concepts', in L. E. Hahn ( ed.), The

Philosophy of Roderick Chisholm, La Salle, IL: Open Court, pp. 217±38.

Wright, C. [1985]: `Facts and Certainty', Proceedings of the British Academy, 71,

pp. 429±72.

Wright, C. [1991]: `Skepticism and Dreaming: Imploding the Demon', Mind , 100,

pp. 87±115.

Wright, C. [2000]: `Cogency and Question-Begging: Some Re¯ections on McKinsey's

Paradox and Putnam's Proof', in E. Villanuova (ed.), Philosophical Issues, 10.

Wright, C., Smith, M., and Macdonald, C. [1998]: Knowing Our Own Minds: Essays inSelf-Knowledge, Oxford: Oxford University Press.

124 James Pryor

Highlights of Recent Epistemology

Documents

Transcript of Highlights of Recent Epistemology